RVD2.7 is an extension of RVD that includes a prior on the position-specific parameters.
The model is The position is j = 1,...,J. The replicate is i = 1,...,N. The model only addresses error/reference positions and does not model individual nucleotide frequencies.
\mu_j | \mu_0, M_0 ~ Beta(\mu_0, M_0) \theta_{ij} | \mu_j, M_j ~ Beta(\mu_j, M_j) r_{ij} | \theta_{ij}, n_{ij} ~ Binomial(\theta_{ij}, n_{ij})
where n_{ij} is the total counts at position j in replicate i.
The error read count at position j in replicate i is modeled by the binomial random variable r_{ij}. The probability of an error at position j in replicate i is \theta_{ij}. The error probability has a prior beta distribution with position-specific rate parameter \mu_j and precision M_j. The position error rate, \mu_j, has as a Beta prior distribution with parameters \mu_0 and M_0. This is to ensure that that error rate is between 0 and 1. The precision parameter M_j has an improper prior. This is useful for situations when there is a significant minor allele.
You can install a command-line version of rvd easily if you have docker installed. Simply run
docker build -t MYTAG .
in the repository directory and it will build a docker image locally with MYTAG. You can then run the code with
docker run -it --rm -v "
pwd":/work MYTAG
Be sure to wrap pwd with backticks to evaluate the command.
This will mount your current directory under /work/ and run rvd27.py from inside that directory.
For example, if you want to run the gibbs command on a file called myfile.dc, you would use
docker run -it --rm -v "
pwd":/work MYTAG gibbs myfile.dc